IDS 702: An Analsis of the Sneaker Resale Market
Nikhil Bhargava
11/21/20
Summary
After analyzing and fitting data scraped from Kicks on Fire and StockX to a multi-linear regression model, this analysis answered the following inferential questions: -Which of the variables appears to be the best covariate for explaining sneaker resale values? -What is the likely range for the difference in resale values for collaborated and non-collaborated sneakers? -Is there any evidence that the association between sentiment and resale values differ by collaborations? -Are there other interesting associations with resale values that are worth mentioning and modeling? Ultimately, it was determined that the log of the resale price of a sneaker was the most important factor in determining the resale value of a sneaker. Additionally,
Introduction
Using data scraped from a sneaker news website, Kicks on Fire, and StockX, a resale marketplace for Sneakers, the aim of this analysis is to understand what factors effect a sneaker’s resale value. The dataset contains over 6000 sneakers released over the past decade with their average resale value after two plus months of initial release. Additional features such as is a sneaker a collaboration, the intended sex for the sneaker release, if a sneaker was retro, or if it is a kids shoe were engineered using regular expressions. The color of a sneaker was also assigned using a vector representation of the colorway string and getting the cosine similiarity of that vector with common colors to determine the “main color” of a sneaker.
Ultimately, this analysis will answer inferential questions such as which variable is the biggest contributor to a sneaker’s resale values, a likely range for differences in resale values for collab and non-collab sneakers, and potential evidence of associations, in particular, between sentiment and resale values differing by collaborations. To answer these questions posed, this analysis will the explore and transform the data, fit multi-linear regression models with appropriate features, and interpret the results.
Exploratory Data Analysis (EDA)
Data Transformation
Before beginning the analysis, the number of null values in the dataset were counted. Although there were no null values found in the predictors, there were 181 rows in the response variable missing values. This was likely due to the scraper unable to find those sneakers on StockX. Ultimately, these rows were removed instead of imputed since they accounted for less than 3% of the entire dataset and it wasn’t apparent if they weren’t on StockX or if they were difficult to find when searching for them. After removing those rows, the final dataset contained 6,164 unique observations and 19 rows, albeit, 5 of them being intermediate columns or text descriptions.
Since one of the assumptions of a linear model is normality, the distribution of the response variable was plotted as a histogram. The distribution of Resale prices were heavily skewed left, and therefore multiple transformation were performed. Ultimately, the log transformation was deemed to have the best resemblance to a normal distribution. Therefore, the log transformation of the response variable was used for the rest of the analysis.
The “wants” column scraped from Kicks on Fire is akin to a like button, and therefore could be used as a proxy for public sentiment towards each sneaker. The distribution of “wants” was heavily skewed left with a wide range of values. Due to this, and since many sneakers had values of equal to zero, “wants” was binned into four groups roughly representing the quartiles of values in the “wants” columns. A sneaker with less than 50 “wants” was considered “Not Popular”, greater than equal to 50 but less than 250 was “Somewhat Popular”, greater than or equal to 250 and less than 1000 was “Very Popular”, and greater than or equal to “1000” was “Extremely Popular”. The figure on the left below visualizes the distribution of each category below.
Data Visualization
Due to our interest in finding evidence between an association between sentiment and resale values by by whether or not a sneaker was collaborated on, the plot on the right visualizes that relationship. This revealed that there does appears to be a difference in resale values among each popularity group when collaborated on versus not. Although the resale distributions for non-collaborated sneakers look similar to the overall distribution of the data set (pictured left), the distributions for sneakers that were collaborated do not. The median resale values for “Somewhat Popular” and “Very Popular” collaborated sneakers appear to be roughly the same, as compared to before. The median value and distribution of extremely popular, collab sneakers, seem to be higher than before as well.
Two additional associations found while performing EDA can be seen in the plot below. The graph on the left shows the log of resale prices versus retail prices split by brand. From this chart, we can make out that some brands, such as Converse, Nike, Air Jordan, and Adidas, all appear to have strong positive correlations between the two variables. Meanwhile, some brands such as Puma and Asics, have relatively flat slopes with no correlation between the two variables. The second graph on the right depicts the log of resale prices versus sneaker color split by collaboration. The resale value medians and distributions for each colors vary a lot more sneakers for sneakers that are collaborated on versus not. In particular, the resale value distributions for brown, purple, red, and white sneakers seem to be a lot higher when collaborated on, while the distribution for gray sneakers seems to decrease. Both of these associations will be tested for statistical significance when modeling.
Model
Model Selection
The final model was selecting based on our findings from the EDA, step-wise model selection, and ANOVA f-tests. When checking model diagnostic plots, it was determined that the log transformation of the retail value for each sneaker would create more randomness among the residuals in the residuals vs fitted plots and was necessary to satisfy the linearity assumption. The rest of the variables were all categorical, and therefore no other transformations to predictors were made. The final model took the form: \(y_i = \beta_0 + \beta_1 Brand + \beta_2 Release Month + \beta_3 Release Day + \beta_4 Sex + \beta_5 Collab + \beta_6 Retro + \beta_7 Kids + \beta_8 Color + \beta_9 BinWants + \beta_{10} log(RetailPrice) + \beta_{11} (BinWants:Collab) + \beta_{12} (log(RetailPrice):Brand) + \beta_{13} (Collab:Color)\)
The final model summary can be seen below. Since the response variable is the log of resale values, it is important to exponentiate the coefficients before interpreting the model.
The model’s intercept is 15.21, which means that, if all variables are set to 0 or their respective baseline, the average resale price of a sneaker would be $15.21. The intercept is also statistically significant as its’ p-value is smaller than 0.05. The adjusted r-squared value for the model is ~0.38, which means about 38% of variability in the dataset can be explained by the variables in our model.
We can also determine that the log of the sneaker’s retail price is the best covariate for explaining sneaker resale values. This is because predictors with higher absolute t-values indicate greater evidence against the null hypothesis and importance in predicting the response variable. The coefficient of this covariate indicates that for every additional increase in the log of the retail price, the resale value, on average, for a sneaker increases by $1.48, holding all other variables constant.
For this analysis, only variables that had p-values lower than the 0.05 level were deemed to be statistically significant. Therefore, brand, release month, release day, retro, “wants” binned, log of retail price, the interaction between brand and the log of retail price, the interaction between collab and “wants” binned, and the interaction between collab and main color were all statistically significant variables.
Additionally, since the interaction between collaborations and “wants” binned was statistically significant, we know that there is evidence of an association between sentiment and resale values different. For the interaction, non-collaborated sneakers and “Not Popular” were the model’s baseline. The statistically significant effects of this interaction can be interpreted as, if all other variables are held constant, on average, for sneakers that were collaborated on and were “Somewhat Popular”, the average resale price of a sneaker increased by $1.13 compared to the model baseline. For the same conditions, on average, sneakers that were collaborated on and were “Extremely Popular”, the average resale price of a sneaker increased by $1.69 compared to the model baseline. Although not statistically significant on the 0.05 level, holding all variables constant, sneakers that were collaborated on and were “Very Popular”, the average resale price of a sneaker increased by $1.12 compared to the model baseline.
| 2.5 % | 97.5 % | |
|---|---|---|
| collabTRUE | 0.9619 | 1.2 |
Next, a likely range for the difference in sneaker resale values for sneakers that were non-collab and collab can be found by getting the 95% confidence interval of the collab variable in the final model, which can be seen above. Based on the 95% confidence interval for the variable collab, if all other variables are held constant, the average resale price of a collab sneaker increases by $0.96 to $1.20 compared non-collab sneakers. It is worth noting that collab was not found to be significant on its own, but was significant in its interaction between sneaker sentiment and color.
| term | estimate | std.error | statistic | p.value |
|---|---|---|---|---|
| (Intercept) | 2.7220249 | 0.1309938 | 20.7797938 | 0.0000000 |
| brandAir Jordan | -1.4736253 | 0.2795425 | -5.2715610 | 0.0000001 |
| brandAsics | 5.8292930 | 3.9994977 | 1.4575063 | 0.1450282 |
| brandConverse | -3.8826690 | 1.9132581 | -2.0293493 | 0.0424661 |
| brandJordan Brand | -1.0193292 | 0.6562504 | -1.5532626 | 0.1204123 |
| brandNew Balance | -3.7749322 | 2.5269023 | -1.4938972 | 0.1352543 |
| brandNike | -1.7853927 | 0.1626169 | -10.9791346 | 0.0000000 |
| brandNike Basketball | -2.1295007 | 0.7247864 | -2.9381078 | 0.0033146 |
| brandNike Running | -0.9793677 | 1.2241818 | -0.8000182 | 0.4237314 |
| brandOther Brands | 0.0420777 | 0.9012705 | 0.0466871 | 0.9627641 |
| brandPuma | 0.0597020 | 1.0646902 | 0.0560746 | 0.9552843 |
| brandReebok | -0.2161552 | 0.8269331 | -0.2613938 | 0.7937977 |
| release_monthFeb | 0.0263528 | 0.0289974 | 0.9087966 | 0.3634935 |
| release_monthMar | 0.0787506 | 0.0289739 | 2.7179847 | 0.0065867 |
| release_monthApr | 0.0246435 | 0.0294154 | 0.8377768 | 0.4021889 |
| release_monthMay | -0.0198294 | 0.0307540 | -0.6447755 | 0.5190969 |
| release_monthJun | 0.0413382 | 0.0297805 | 1.3880963 | 0.1651585 |
| release_monthJul | 0.0167872 | 0.0303453 | 0.5532063 | 0.5801424 |
| release_monthAug | 0.0262351 | 0.0299913 | 0.8747593 | 0.3817394 |
| release_monthSep | 0.0685826 | 0.0309396 | 2.2166588 | 0.0266832 |
| release_monthOct | 0.0401937 | 0.0316186 | 1.2712069 | 0.2037036 |
| release_monthNov | 0.0872633 | 0.0299581 | 2.9128457 | 0.0035945 |
| release_monthDec | 0.0826664 | 0.0303689 | 2.7220703 | 0.0065059 |
| release_dowMonday | 0.1008594 | 0.0351539 | 2.8690810 | 0.0041308 |
| release_dowTuesday | 0.0225498 | 0.0327940 | 0.6876176 | 0.4917198 |
| release_dowWednesday | 0.0251268 | 0.0300584 | 0.8359345 | 0.4032245 |
| release_dowThursday | -0.0109732 | 0.0266154 | -0.4122877 | 0.6801431 |
| release_dowFriday | 0.0091321 | 0.0278366 | 0.3280599 | 0.7428777 |
| release_dowSaturday | 0.0359227 | 0.0270957 | 1.3257686 | 0.1849660 |
| wmnsTRUE | 0.0262820 | 0.0160629 | 1.6361880 | 0.1018519 |
| collabTRUE | 0.0716579 | 0.0563916 | 1.2707185 | 0.2038773 |
| retroTRUE | 0.1620268 | 0.0323543 | 5.0078971 | 0.0000006 |
| kidsTRUE | -0.0153924 | 0.0228705 | -0.6730217 | 0.5009590 |
| main_colorBlue | 0.0062015 | 0.0170300 | 0.3641490 | 0.7157594 |
| main_colorBrown | -0.0273718 | 0.0381856 | -0.7168096 | 0.4735190 |
| main_colorGray | -0.0098409 | 0.0280986 | -0.3502283 | 0.7261795 |
| main_colorGreen | -0.0021702 | 0.0386206 | -0.0561918 | 0.9551909 |
| main_colorMulti-Color | 0.0833139 | 0.0425781 | 1.9567304 | 0.0504248 |
| main_colorOrange | -0.0628058 | 0.0384314 | -1.6342321 | 0.1022618 |
| main_colorPurple | -0.0292959 | 0.0369776 | -0.7922620 | 0.4282388 |
| main_colorRed | 0.0083963 | 0.0238334 | 0.3522905 | 0.7246326 |
| main_colorWhite | 0.0147409 | 0.0150038 | 0.9824734 | 0.3259057 |
| main_colorYellow | -0.0972134 | 0.0730376 | -1.3310053 | 0.1832371 |
| bin_wantsSomewhat Popular | -0.0248405 | 0.0176301 | -1.4089882 | 0.1588897 |
| bin_wantsVery Popular | 0.0494566 | 0.0183010 | 2.7023988 | 0.0069032 |
| bin_wantsExtremely Popular | 0.1771991 | 0.0199648 | 8.8755807 | 0.0000000 |
| log_retail_price | 0.3921911 | 0.0263512 | 14.8832570 | 0.0000000 |
| collabTRUE:bin_wantsSomewhat Popular | 0.1199308 | 0.0592720 | 2.0233968 | 0.0430758 |
| collabTRUE:bin_wantsVery Popular | 0.1120288 | 0.0585095 | 1.9147114 | 0.0555761 |
| collabTRUE:bin_wantsExtremely Popular | 0.5248961 | 0.0615767 | 8.5242597 | 0.0000000 |
| brandAir Jordan:log_retail_price | 0.3238057 | 0.0553798 | 5.8470028 | 0.0000000 |
| brandAsics:log_retail_price | -1.1940970 | 0.7962466 | -1.4996572 | 0.1337550 |
| brandConverse:log_retail_price | 0.8384674 | 0.4181146 | 2.0053531 | 0.0449695 |
| brandJordan Brand:log_retail_price | 0.1898603 | 0.1341948 | 1.4148115 | 0.1571749 |
| brandNew Balance:log_retail_price | 0.7353419 | 0.4998966 | 1.4709880 | 0.1413460 |
| brandNike:log_retail_price | 0.3736445 | 0.0333368 | 11.2081817 | 0.0000000 |
| brandNike Basketball:log_retail_price | 0.4352096 | 0.1406644 | 3.0939576 | 0.0019840 |
| brandNike Running:log_retail_price | 0.2321527 | 0.2520508 | 0.9210552 | 0.3570581 |
| brandOther Brands:log_retail_price | -0.0127406 | 0.1849630 | -0.0688816 | 0.9450861 |
| brandPuma:log_retail_price | -0.0637214 | 0.2202884 | -0.2892636 | 0.7723895 |
| brandReebok:log_retail_price | 0.0502490 | 0.1679321 | 0.2992225 | 0.7647805 |
| collabTRUE:main_colorBlue | -0.0863874 | 0.0534528 | -1.6161427 | 0.1061152 |
| collabTRUE:main_colorBrown | -0.0258520 | 0.0913443 | -0.2830168 | 0.7771736 |
| collabTRUE:main_colorGray | -0.3347995 | 0.1242360 | -2.6948667 | 0.0070610 |
| collabTRUE:main_colorGreen | 0.1345769 | 0.1236725 | 1.0881715 | 0.2765624 |
| collabTRUE:main_colorMulti-Color | -0.0872436 | 0.1087129 | -0.8025135 | 0.4222873 |
| collabTRUE:main_colorOrange | 0.1608573 | 0.0918387 | 1.7515207 | 0.0799066 |
| collabTRUE:main_colorPurple | 0.0550746 | 0.1054543 | 0.5222600 | 0.6015083 |
| collabTRUE:main_colorRed | 0.0709415 | 0.0834196 | 0.8504179 | 0.3951262 |
| collabTRUE:main_colorWhite | 0.1100795 | 0.0470256 | 2.3408402 | 0.0192725 |
| collabTRUE:main_colorYellow | 0.0453065 | 0.1552968 | 0.2917414 | 0.7704943 |
Model Assesment
The model diagnostics were the plotted below. Looking at the residuals vs. fitted plot, the model does seem to satisfy the independence and equal variance assumptions as the data points are randomly scattered throughout the plot and the red line indicating trends is straight.
The Normal Q-Q plot indicates whether or not the assumption of Normality is violated by the model. Looking at the plot, the points roughly lie on the dashed line drawn, until the right most extremity. This is likely due to the long right tail of the response variable distribution. Therefore, the normality assumption does appear to be satisfied.
Potential outliers, leverage points, or highly influential points can all be identified using the Residuals vs. Leverage plot below. From the Cook’s distance lines, there are no highly influential data points as there are not values with a Cook’s distance value greater than 0.5 or 1. This meant no data points were highly influential and did not need to be removed from our analysis. High leverage points were calculated using the formula \(2*(p+1)/n\). From this threshold there were quite a few high leverage points. Outliers could also be identified from the Residuals vs. Leverage plot below by looking at the y-values of the data points in the graph. The distribution of data points appear to follow a normal distribution and can be found between 3 and -3 standard deviations. However, there are quite a few data points that are outside those metrics.
Finally, the Variance Inflation Factor (VIF) score for each variable in the model was calculated to determine how much multicollinearity between the variables in the model was inflating the variance of the regression coefficient for each variable. However, since the only continuous variable was the log of the retail price, that was the only VIF score worth noting. The log of the retail price had a VIF score of 3.31, and therefore we can safely say it’s not correlated with other variables because it’s lower than the high correlation threshold of 10.
Conclusion
In conclusion, our analysis determined that the best covariate for explaining sneaker resale values was the log of the resale price. In general, the log of the resale price was positively correlated with resale values. Additionally, there was an association found between sentiment and resale values differing by collaborations. As discussed earlier, sneakers that were collaborated on and were “Somewhat Popular” or “Extremely Popular” usually had higher resale values than “Not Popular” and non-collab sneakers. There were also associations found between the log of the resale price and resale values differing by Brand, as well as the main color and resale values of sneakers differing by collaborations. Although collaborations were not found to be statistically significant, the likely difference in resale values for collaborated and non-collab sneakers was found to be between an increase in $0.96 to $1.20 for collab sneakers versus not.
Overall, there were a few limitations of this analysis. First, the dataset built from Kicks on Fire is not an exhaustive list of every sneaker ever released and is general biased towards talking about more “trendy” sneakers. Therefore, some brands may appear to only have sneakers with high resale values, such as Converse. Second, determining the main color for a sneaker is tough job for a human to label, let alone a computer. Although the labels seemed generally accurate, the main color of a sneaker is hard to quantify and up for debate. Third, although the response variable is an average resale price of sneakers, According to StockX there has been found to be a range of average prices between different shoe sizes of the same sneaker, which is something the model doesn’t take into account.
Appendix
III. EDA Distributions of response variable 
Distributions of retail price 
EDA for log resale vs brand 
EDA for log resale vs main color 
EDA for log resale vs release month 
EDA for log resale vs release day 
EDA for log resale vs binned “wants” 
EDA for log resale vs log retail 
Model
##
## Call:
## lm(formula = log_resale_price ~ brand + release_month + release_dow +
## wmns + collab + retro + kids + main_color + bin_wants + retail_price,
## data = sneakers)
##
## Residuals:
## Min 1Q Median 3Q Max
## -2.5217 -0.2455 -0.0107 0.2084 3.7741
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 4.0792824 0.0385592 105.793 < 2e-16 ***
## brandAir Jordan 0.1403510 0.0216334 6.488 9.40e-11 ***
## brandAsics -0.1597796 0.0901347 -1.773 0.076332 .
## brandConverse -0.0131283 0.0877316 -0.150 0.881052
## brandJordan Brand -0.0834422 0.0397535 -2.099 0.035859 *
## brandNew Balance -0.0980314 0.1314188 -0.746 0.455728
## brandNike 0.0523806 0.0144458 3.626 0.000290 ***
## brandNike Basketball 0.0401751 0.0404242 0.994 0.320342
## brandNike Running 0.1332712 0.0776873 1.715 0.086308 .
## brandOther Brands -0.0214001 0.0455201 -0.470 0.638283
## brandPuma -0.2229885 0.0594071 -3.754 0.000176 ***
## brandReebok 0.0217633 0.0463381 0.470 0.638612
## release_monthFeb 0.0214823 0.0297217 0.723 0.469842
## release_monthMar 0.0553728 0.0296907 1.865 0.062231 .
## release_monthApr 0.0200584 0.0301602 0.665 0.506036
## release_monthMay -0.0313537 0.0315031 -0.995 0.319650
## release_monthJun 0.0292981 0.0304838 0.961 0.336539
## release_monthJul -0.0024515 0.0310468 -0.079 0.937066
## release_monthAug 0.0223310 0.0307258 0.727 0.467386
## release_monthSep 0.0661192 0.0317232 2.084 0.037178 *
## release_monthOct 0.0424801 0.0324080 1.311 0.189978
## release_monthNov 0.0910831 0.0306954 2.967 0.003016 **
## release_monthDec 0.0709438 0.0310805 2.283 0.022489 *
## release_dowMonday 0.1773772 0.0358024 4.954 7.45e-07 ***
## release_dowTuesday 0.0686351 0.0332976 2.061 0.039320 *
## release_dowWednesday 0.0626654 0.0306779 2.043 0.041126 *
## release_dowThursday 0.0471950 0.0270069 1.748 0.080598 .
## release_dowFriday 0.0443193 0.0283175 1.565 0.117614
## release_dowSaturday 0.0783300 0.0276095 2.837 0.004568 **
## wmnsTRUE 0.0331839 0.0164615 2.016 0.043860 *
## collabTRUE 0.2660243 0.0196380 13.546 < 2e-16 ***
## retroTRUE 0.1481206 0.0330190 4.486 7.39e-06 ***
## kidsTRUE -0.0635748 0.0228878 -2.778 0.005492 **
## main_colorBlue 0.0018315 0.0165285 0.111 0.911770
## main_colorBrown -0.0623182 0.0355700 -1.752 0.079825 .
## main_colorGray -0.0096130 0.0280403 -0.343 0.731741
## main_colorGreen 0.0159876 0.0376303 0.425 0.670954
## main_colorMulti-Color 0.0612667 0.0401180 1.527 0.126772
## main_colorOrange -0.0287045 0.0356582 -0.805 0.420856
## main_colorPurple -0.0219321 0.0354176 -0.619 0.535780
## main_colorRed 0.0214855 0.0234209 0.917 0.358989
## main_colorWhite 0.0266488 0.0145803 1.828 0.067638 .
## main_colorYellow -0.0998767 0.0659459 -1.515 0.129944
## bin_wantsSomewhat Popular 0.0126271 0.0170421 0.741 0.458763
## bin_wantsVery Popular 0.1147260 0.0176036 6.517 7.74e-11 ***
## bin_wantsExtremely Popular 0.3037893 0.0192095 15.815 < 2e-16 ***
## retail_price 0.0033194 0.0001017 32.633 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.4298 on 6117 degrees of freedom
## Multiple R-squared: 0.3494, Adjusted R-squared: 0.3445
## F-statistic: 71.4 on 46 and 6117 DF, p-value: < 2.2e-16
##
## Call:
## lm(formula = log_resale_price ~ brand + release_month + release_dow +
## wmns + collab + retro + kids + main_color + bin_wants + log_retail_price,
## data = sneakers)
##
## Residuals:
## Min 1Q Median 3Q Max
## -1.7783 -0.2377 -0.0097 0.1857 3.9719
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1.686911 0.089893 18.766 < 2e-16 ***
## brandAir Jordan 0.113248 0.021603 5.242 1.64e-07 ***
## brandAsics -0.200110 0.089656 -2.232 0.025652 *
## brandConverse 0.020547 0.087356 0.235 0.814054
## brandJordan Brand -0.098477 0.039542 -2.490 0.012786 *
## brandNew Balance -0.123857 0.130756 -0.947 0.343554
## brandNike 0.040374 0.014372 2.809 0.004982 **
## brandNike Basketball 0.021146 0.040244 0.525 0.599295
## brandNike Running 0.140815 0.077306 1.822 0.068574 .
## brandOther Brands -0.027090 0.045276 -0.598 0.549648
## brandPuma -0.218746 0.059111 -3.701 0.000217 ***
## brandReebok 0.004719 0.046078 0.102 0.918427
## release_monthFeb 0.027915 0.029574 0.944 0.345255
## release_monthMar 0.072354 0.029539 2.449 0.014334 *
## release_monthApr 0.027865 0.030012 0.928 0.353194
## release_monthMay -0.027424 0.031348 -0.875 0.381706
## release_monthJun 0.041084 0.030342 1.354 0.175780
## release_monthJul 0.009847 0.030901 0.319 0.750007
## release_monthAug 0.023808 0.030573 0.779 0.436170
## release_monthSep 0.066245 0.031565 2.099 0.035884 *
## release_monthOct 0.041892 0.032246 1.299 0.193949
## release_monthNov 0.080164 0.030542 2.625 0.008693 **
## release_monthDec 0.070657 0.030926 2.285 0.022362 *
## release_dowMonday 0.148086 0.035665 4.152 3.34e-05 ***
## release_dowTuesday 0.053595 0.033142 1.617 0.105901
## release_dowWednesday 0.038266 0.030548 1.253 0.210373
## release_dowThursday 0.020888 0.026910 0.776 0.437654
## release_dowFriday 0.030982 0.028200 1.099 0.271962
## release_dowSaturday 0.049302 0.027518 1.792 0.073235 .
## wmnsTRUE 0.029162 0.016375 1.781 0.074975 .
## collabTRUE 0.272387 0.019503 13.966 < 2e-16 ***
## retroTRUE 0.147183 0.032850 4.481 7.58e-06 ***
## kidsTRUE -0.015373 0.023028 -0.668 0.504436
## main_colorBlue -0.002759 0.016447 -0.168 0.866763
## main_colorBrown -0.044499 0.035362 -1.258 0.208307
## main_colorGray -0.028693 0.027914 -1.028 0.304033
## main_colorGreen 0.008407 0.037442 0.225 0.822342
## main_colorMulti-Color 0.062279 0.039917 1.560 0.118761
## main_colorOrange -0.031986 0.035482 -0.901 0.367373
## main_colorPurple -0.024294 0.035242 -0.689 0.490634
## main_colorRed 0.020981 0.023304 0.900 0.367993
## main_colorWhite 0.023514 0.014504 1.621 0.105040
## main_colorYellow -0.110867 0.065620 -1.690 0.091169 .
## bin_wantsSomewhat Popular -0.031493 0.017210 -1.830 0.067315 .
## bin_wantsVery Popular 0.053276 0.017934 2.971 0.002982 **
## bin_wantsExtremely Popular 0.242763 0.019599 12.387 < 2e-16 ***
## log_retail_price 0.598942 0.017763 33.719 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.4276 on 6117 degrees of freedom
## Multiple R-squared: 0.3558, Adjusted R-squared: 0.351
## F-statistic: 73.46 on 46 and 6117 DF, p-value: < 2.2e-16
##
## Call:
## lm(formula = log_resale_price ~ brand + release_month + release_dow +
## wmns + collab + retro + kids + main_color + bin_wants + log_retail_price +
## bin_wants:collab + log_retail_price:brand + collab:main_color,
## data = sneakers)
##
## Residuals:
## Min 1Q Median 3Q Max
## -1.6802 -0.2265 0.0077 0.1738 3.7662
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 2.722025 0.130994 20.780 < 2e-16 ***
## brandAir Jordan -1.473625 0.279543 -5.272 1.40e-07 ***
## brandAsics 5.829293 3.999498 1.458 0.14503
## brandConverse -3.882669 1.913258 -2.029 0.04247 *
## brandJordan Brand -1.019329 0.656250 -1.553 0.12041
## brandNew Balance -3.774932 2.526902 -1.494 0.13525
## brandNike -1.785393 0.162617 -10.979 < 2e-16 ***
## brandNike Basketball -2.129501 0.724786 -2.938 0.00331 **
## brandNike Running -0.979368 1.224182 -0.800 0.42373
## brandOther Brands 0.042078 0.901270 0.047 0.96276
## brandPuma 0.059702 1.064690 0.056 0.95528
## brandReebok -0.216155 0.826933 -0.261 0.79380
## release_monthFeb 0.026353 0.028997 0.909 0.36349
## release_monthMar 0.078751 0.028974 2.718 0.00659 **
## release_monthApr 0.024644 0.029415 0.838 0.40219
## release_monthMay -0.019829 0.030754 -0.645 0.51910
## release_monthJun 0.041338 0.029781 1.388 0.16516
## release_monthJul 0.016787 0.030345 0.553 0.58014
## release_monthAug 0.026235 0.029991 0.875 0.38174
## release_monthSep 0.068583 0.030940 2.217 0.02668 *
## release_monthOct 0.040194 0.031619 1.271 0.20370
## release_monthNov 0.087263 0.029958 2.913 0.00359 **
## release_monthDec 0.082666 0.030369 2.722 0.00651 **
## release_dowMonday 0.100859 0.035154 2.869 0.00413 **
## release_dowTuesday 0.022550 0.032794 0.688 0.49172
## release_dowWednesday 0.025127 0.030058 0.836 0.40322
## release_dowThursday -0.010973 0.026615 -0.412 0.68014
## release_dowFriday 0.009132 0.027837 0.328 0.74288
## release_dowSaturday 0.035923 0.027096 1.326 0.18497
## wmnsTRUE 0.026282 0.016063 1.636 0.10185
## collabTRUE 0.071658 0.056392 1.271 0.20388
## retroTRUE 0.162027 0.032354 5.008 5.66e-07 ***
## kidsTRUE -0.015392 0.022871 -0.673 0.50096
## main_colorBlue 0.006201 0.017030 0.364 0.71576
## main_colorBrown -0.027372 0.038186 -0.717 0.47352
## main_colorGray -0.009841 0.028099 -0.350 0.72618
## main_colorGreen -0.002170 0.038621 -0.056 0.95519
## main_colorMulti-Color 0.083314 0.042578 1.957 0.05042 .
## main_colorOrange -0.062806 0.038431 -1.634 0.10226
## main_colorPurple -0.029296 0.036978 -0.792 0.42824
## main_colorRed 0.008396 0.023833 0.352 0.72463
## main_colorWhite 0.014741 0.015004 0.982 0.32591
## main_colorYellow -0.097213 0.073038 -1.331 0.18324
## bin_wantsSomewhat Popular -0.024841 0.017630 -1.409 0.15889
## bin_wantsVery Popular 0.049457 0.018301 2.702 0.00690 **
## bin_wantsExtremely Popular 0.177199 0.019965 8.876 < 2e-16 ***
## log_retail_price 0.392191 0.026351 14.883 < 2e-16 ***
## collabTRUE:bin_wantsSomewhat Popular 0.119931 0.059272 2.023 0.04308 *
## collabTRUE:bin_wantsVery Popular 0.112029 0.058509 1.915 0.05558 .
## collabTRUE:bin_wantsExtremely Popular 0.524896 0.061577 8.524 < 2e-16 ***
## brandAir Jordan:log_retail_price 0.323806 0.055380 5.847 5.26e-09 ***
## brandAsics:log_retail_price -1.194097 0.796247 -1.500 0.13376
## brandConverse:log_retail_price 0.838467 0.418115 2.005 0.04497 *
## brandJordan Brand:log_retail_price 0.189860 0.134195 1.415 0.15717
## brandNew Balance:log_retail_price 0.735342 0.499897 1.471 0.14135
## brandNike:log_retail_price 0.373645 0.033337 11.208 < 2e-16 ***
## brandNike Basketball:log_retail_price 0.435210 0.140664 3.094 0.00198 **
## brandNike Running:log_retail_price 0.232153 0.252051 0.921 0.35706
## brandOther Brands:log_retail_price -0.012741 0.184963 -0.069 0.94509
## brandPuma:log_retail_price -0.063721 0.220288 -0.289 0.77239
## brandReebok:log_retail_price 0.050249 0.167932 0.299 0.76478
## collabTRUE:main_colorBlue -0.086387 0.053453 -1.616 0.10612
## collabTRUE:main_colorBrown -0.025852 0.091344 -0.283 0.77717
## collabTRUE:main_colorGray -0.334800 0.124236 -2.695 0.00706 **
## collabTRUE:main_colorGreen 0.134577 0.123672 1.088 0.27656
## collabTRUE:main_colorMulti-Color -0.087244 0.108713 -0.803 0.42229
## collabTRUE:main_colorOrange 0.160857 0.091839 1.752 0.07991 .
## collabTRUE:main_colorPurple 0.055075 0.105454 0.522 0.60151
## collabTRUE:main_colorRed 0.070942 0.083420 0.850 0.39513
## collabTRUE:main_colorWhite 0.110080 0.047026 2.341 0.01927 *
## collabTRUE:main_colorYellow 0.045307 0.155297 0.292 0.77049
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.4186 on 6093 degrees of freedom
## Multiple R-squared: 0.3851, Adjusted R-squared: 0.378
## F-statistic: 54.51 on 70 and 6093 DF, p-value: < 2.2e-16
## Analysis of Variance Table
##
## Model 1: log_resale_price ~ brand + release_month + release_dow + wmns +
## collab + retro + kids + main_color + bin_wants + log_retail_price
## Model 2: log_resale_price ~ brand + release_month + release_dow + wmns +
## collab + retro + kids + main_color + bin_wants + log_retail_price +
## bin_wants:collab + log_retail_price:brand + collab:main_color
## Res.Df RSS Df Sum of Sq F Pr(>F)
## 1 6117 1118.6
## 2 6093 1067.8 24 50.809 12.08 < 2.2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1##
## Call:
## lm(formula = log_resale_price ~ collab + log_retail_price + bin_wants +
## brand + retro + release_month + release_dow + wmns + collab:bin_wants +
## log_retail_price:brand, data = sneakers)
##
## Residuals:
## Min 1Q Median 3Q Max
## -1.6163 -0.2313 0.0087 0.1751 3.7529
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 2.733536 0.129370 21.130 < 2e-16 ***
## collabTRUE 0.114659 0.049252 2.328 0.01994 *
## log_retail_price 0.389646 0.026034 14.967 < 2e-16 ***
## bin_wantsSomewhat Popular -0.023223 0.017589 -1.320 0.18677
## bin_wantsVery Popular 0.052436 0.018223 2.878 0.00402 **
## bin_wantsExtremely Popular 0.180985 0.019853 9.116 < 2e-16 ***
## brandAir Jordan -1.500470 0.276037 -5.436 5.67e-08 ***
## brandAsics 5.534953 3.993674 1.386 0.16582
## brandConverse -4.169570 1.911670 -2.181 0.02921 *
## brandJordan Brand -1.045891 0.656973 -1.592 0.11144
## brandNew Balance -4.481266 2.523946 -1.775 0.07587 .
## brandNike -1.793524 0.162459 -11.040 < 2e-16 ***
## brandNike Basketball -2.197946 0.724823 -3.032 0.00244 **
## brandNike Running -0.983794 1.224816 -0.803 0.42188
## brandOther Brands 0.041601 0.897408 0.046 0.96303
## brandPuma 0.375869 1.056407 0.356 0.72200
## brandReebok -0.319454 0.824582 -0.387 0.69846
## retroTRUE 0.163526 0.032267 5.068 4.14e-07 ***
## release_monthFeb 0.027853 0.028986 0.961 0.33663
## release_monthMar 0.077363 0.028946 2.673 0.00755 **
## release_monthApr 0.026515 0.029396 0.902 0.36710
## release_monthMay -0.019191 0.030739 -0.624 0.53242
## release_monthJun 0.041827 0.029781 1.404 0.16022
## release_monthJul 0.017751 0.030354 0.585 0.55872
## release_monthAug 0.025042 0.030003 0.835 0.40394
## release_monthSep 0.069895 0.030961 2.258 0.02401 *
## release_monthOct 0.039128 0.031611 1.238 0.21584
## release_monthNov 0.085086 0.029963 2.840 0.00453 **
## release_monthDec 0.082409 0.030361 2.714 0.00666 **
## release_dowMonday 0.105823 0.035145 3.011 0.00261 **
## release_dowTuesday 0.022403 0.032776 0.684 0.49430
## release_dowWednesday 0.021926 0.030088 0.729 0.46620
## release_dowThursday -0.009759 0.026616 -0.367 0.71387
## release_dowFriday 0.007491 0.027830 0.269 0.78782
## release_dowSaturday 0.034916 0.027094 1.289 0.19755
## wmnsTRUE 0.026267 0.015809 1.662 0.09665 .
## collabTRUE:bin_wantsSomewhat Popular 0.093380 0.058934 1.584 0.11314
## collabTRUE:bin_wantsVery Popular 0.084190 0.058063 1.450 0.14712
## collabTRUE:bin_wantsExtremely Popular 0.508422 0.061075 8.325 < 2e-16 ***
## log_retail_price:brandAir Jordan 0.328881 0.054844 5.997 2.13e-09 ***
## log_retail_price:brandAsics -1.136952 0.795104 -1.430 0.15279
## log_retail_price:brandConverse 0.906474 0.417630 2.171 0.03001 *
## log_retail_price:brandJordan Brand 0.195821 0.134331 1.458 0.14496
## log_retail_price:brandNew Balance 0.874204 0.499361 1.751 0.08006 .
## log_retail_price:brandNike 0.375747 0.033291 11.287 < 2e-16 ***
## log_retail_price:brandNike Basketball 0.447987 0.140705 3.184 0.00146 **
## log_retail_price:brandNike Running 0.232884 0.252175 0.923 0.35578
## log_retail_price:brandOther Brands -0.012604 0.184143 -0.068 0.94543
## log_retail_price:brandPuma -0.132029 0.218608 -0.604 0.54590
## log_retail_price:brandReebok 0.068942 0.167460 0.412 0.68058
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.4194 on 6114 degrees of freedom
## Multiple R-squared: 0.3807, Adjusted R-squared: 0.3757
## F-statistic: 76.7 on 49 and 6114 DF, p-value: < 2.2e-16##
## Call:
## lm(formula = log_resale_price ~ collab + log_retail_price + bin_wants +
## retro + collab:bin_wants, data = sneakers)
##
## Residuals:
## Min 1Q Median 3Q Max
## -1.7486 -0.2353 0.0000 0.1787 3.9516
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1.63925 0.07945 20.632 < 2e-16 ***
## collabTRUE 0.07159 0.04964 1.442 0.149312
## log_retail_price 0.63395 0.01698 37.341 < 2e-16 ***
## bin_wantsSomewhat Popular -0.03505 0.01701 -2.061 0.039390 *
## bin_wantsVery Popular 0.05897 0.01714 3.440 0.000586 ***
## bin_wantsExtremely Popular 0.21847 0.01767 12.364 < 2e-16 ***
## retroTRUE 0.19876 0.03155 6.300 3.19e-10 ***
## collabTRUE:bin_wantsSomewhat Popular 0.07812 0.05956 1.312 0.189690
## collabTRUE:bin_wantsVery Popular 0.07705 0.05829 1.322 0.186251
## collabTRUE:bin_wantsExtremely Popular 0.50417 0.06165 8.178 3.49e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.4277 on 6154 degrees of freedom
## Multiple R-squared: 0.3517, Adjusted R-squared: 0.3508
## F-statistic: 371 on 9 and 6154 DF, p-value: < 2.2e-16## Analysis of Variance Table
##
## Model 1: log_resale_price ~ collab + log_retail_price + bin_wants + retro +
## collab:bin_wants
## Model 2: log_resale_price ~ collab + log_retail_price + bin_wants + brand +
## retro + release_month + release_dow + wmns + collab:bin_wants +
## log_retail_price:brand
## Res.Df RSS Df Sum of Sq F Pr(>F)
## 1 6154 1125.7
## 2 6114 1075.4 40 50.319 7.152 < 2.2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1## Analysis of Variance Table
##
## Model 1: log_resale_price ~ collab + log_retail_price + bin_wants + brand +
## retro + release_month + release_dow + wmns + collab:bin_wants +
## log_retail_price:brand
## Model 2: log_resale_price ~ brand + release_month + release_dow + wmns +
## collab + retro + kids + main_color + bin_wants + log_retail_price +
## bin_wants:collab + log_retail_price:brand + collab:main_color
## Res.Df RSS Df Sum of Sq F Pr(>F)
## 1 6114 1075.4
## 2 6093 1067.8 21 7.6161 2.0695 0.002817 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1## 2.5 % 97.5 %
## (Intercept) 1.176620e+01 1.966458e+01
## brandAir Jordan 1.324398e-01 3.962844e-01
## brandAsics 1.338366e-01 8.643403e+05
## brandConverse 4.840458e-04 8.763348e-01
## brandJordan Brand 9.967784e-02 1.306241e+00
## brandNew Balance 1.618964e-04 3.250112e+00
## brandNike 1.219457e-01 2.307072e-01
## brandNike Basketball 2.871492e-02 4.923019e-01
## brandNike Running 3.407481e-02 4.139029e+00
## brandOther Brands 1.782210e-01 6.103646e+00
## brandPuma 1.316686e-01 8.558040e+00
## brandReebok 1.592570e-01 4.075225e+00
## release_monthFeb 9.699678e-01 1.086757e+00
## release_monthMar 1.022194e+00 1.145166e+00
## release_monthApr 9.675182e-01 1.085790e+00
## release_monthMay 9.230073e-01 1.041289e+00
## release_monthJun 9.831023e-01 1.104860e+00
## release_monthJul 9.581985e-01 1.079259e+00
## release_monthAug 9.679660e-01 1.088748e+00
## release_monthSep 1.007962e+00 1.137958e+00
## release_monthOct 9.784459e-01 1.107580e+00
## release_monthNov 1.028946e+00 1.157187e+00
## release_monthDec 1.023402e+00 1.152807e+00
## release_dowMonday 1.032461e+00 1.185036e+00
## release_dowTuesday 9.591209e-01 1.090720e+00
## release_dowWednesday 9.667666e-01 1.087685e+00
## release_dowThursday 9.388039e-01 1.042063e+00
## release_dowFriday 9.555792e-01 1.065774e+00
## release_dowSaturday 9.829524e-01 1.093124e+00
## wmnsTRUE 9.948065e-01 1.059472e+00
## collabTRUE 9.618569e-01 1.199861e+00
## retroTRUE 1.103626e+00 1.252890e+00
## kidsTRUE 9.415511e-01 1.029880e+00
## main_colorBlue 9.731827e-01 1.040380e+00
## main_colorBrown 9.028227e-01 1.048631e+00
## main_colorGray 9.371387e-01 1.046281e+00
## main_colorGreen 9.250752e-01 1.076311e+00
## main_colorMulti-Color 9.998458e-01 1.181497e+00
## main_colorOrange 8.709724e-01 1.012612e+00
## main_colorPurple 9.032237e-01 1.044140e+00
## main_colorRed 9.623995e-01 1.056666e+00
## main_colorWhite 9.854352e-01 1.045143e+00
## main_colorYellow 7.863189e-01 1.047039e+00
## bin_wantsSomewhat Popular 9.423282e-01 1.009768e+00
## bin_wantsVery Popular 1.013673e+00 1.089080e+00
## bin_wantsExtremely Popular 1.148046e+00 1.241521e+00
## log_retail_price 1.405697e+00 1.558695e+00
## collabTRUE:bin_wantsSomewhat Popular 1.003744e+00 1.266332e+00
## collabTRUE:bin_wantsVery Popular 9.973331e-01 1.254489e+00
## collabTRUE:bin_wantsExtremely Popular 1.498079e+00 1.907147e+00
## brandAir Jordan:log_retail_price 1.240162e+00 1.540905e+00
## brandAsics:log_retail_price 6.360763e-02 1.443149e+00
## brandConverse:log_retail_price 1.018993e+00 5.249431e+00
## brandJordan Brand:log_retail_price 9.294067e-01 1.572913e+00
## brandNew Balance:log_retail_price 7.829926e-01 5.558431e+00
## brandNike:log_retail_price 1.361099e+00 1.551150e+00
## brandNike Basketball:log_retail_price 1.172875e+00 2.035948e+00
## brandNike Running:log_retail_price 7.695449e-01 2.067337e+00
## brandOther Brands:log_retail_price 6.870602e-01 1.418858e+00
## brandPuma:log_retail_price 6.092268e-01 1.445018e+00
## brandReebok:log_retail_price 7.565723e-01 1.461488e+00
## collabTRUE:main_colorBlue 8.259890e-01 1.018569e+00
## collabTRUE:main_colorBrown 8.147132e-01 1.165576e+00
## collabTRUE:main_colorGray 5.608252e-01 9.127867e-01
## collabTRUE:main_colorGreen 8.977489e-01 1.457931e+00
## collabTRUE:main_colorMulti-Color 7.405521e-01 1.134137e+00
## collabTRUE:main_colorOrange 9.810038e-01 1.406204e+00
## collabTRUE:main_colorPurple 8.592863e-01 1.299269e+00
## collabTRUE:main_colorRed 9.115668e-01 1.264243e+00
## collabTRUE:main_colorWhite 1.018054e+00 1.224174e+00
## collabTRUE:main_colorYellow 7.717226e-01 1.418703e+00## brandAir Jordan brandAsics
## 353.059584 2182.109071
## brandConverse brandJordan Brand
## 520.080647 319.790886
## brandNew Balance brandNike
## 400.075956 232.428448
## brandNike Basketball brandNike Running
## 401.539087 272.226211
## brandOther Brands brandPuma
## 451.504644 365.287879
## brandReebok release_monthFeb
## 372.539592 2.698560
## release_monthMar release_monthApr
## 2.758851 3.063686
## release_monthMay release_monthJun
## 2.352315 2.510071
## release_monthJul release_monthAug
## 2.383284 2.524564
## release_monthSep release_monthOct
## 2.231865 2.108956
## release_monthNov release_monthDec
## 2.391471 2.320492
## release_dowMonday release_dowTuesday
## 1.814133 2.268680
## release_dowWednesday release_dowThursday
## 2.704698 4.628991
## release_dowFriday release_dowSaturday
## 4.155572 5.407437
## wmnsTRUE collabTRUE
## 1.186176 10.522385
## retroTRUE kidsTRUE
## 1.116743 1.227197
## main_colorBlue main_colorBrown
## 1.441675 1.304606
## main_colorGray main_colorGreen
## 1.154905 1.172617
## main_colorMulti-Color main_colorOrange
## 1.256948 1.305476
## main_colorPurple main_colorRed
## 1.230762 1.246164
## main_colorWhite main_colorYellow
## 1.480763 1.329789
## bin_wantsSomewhat Popular bin_wantsVery Popular
## 1.920047 2.200202
## bin_wantsExtremely Popular log_retail_price
## 2.665832 3.313898
## collabTRUE:bin_wantsSomewhat Popular collabTRUE:bin_wantsVery Popular
## 3.729120 4.252918
## collabTRUE:bin_wantsExtremely Popular brandAir Jordan:log_retail_price
## 3.187039 357.990262
## brandAsics:log_retail_price brandConverse:log_retail_price
## 2182.882839 520.426173
## brandJordan Brand:log_retail_price brandNew Balance:log_retail_price
## 319.826771 400.074258
## brandNike:log_retail_price brandNike Basketball:log_retail_price
## 238.020203 402.871264
## brandNike Running:log_retail_price brandOther Brands:log_retail_price
## 272.192130 451.642426
## brandPuma:log_retail_price brandReebok:log_retail_price
## 364.962463 372.424046
## collabTRUE:main_colorBlue collabTRUE:main_colorBrown
## 1.839961 1.421393
## collabTRUE:main_colorGray collabTRUE:main_colorGreen
## 1.142537 1.219088
## collabTRUE:main_colorMulti-Color collabTRUE:main_colorOrange
## 1.344403 1.436819
## collabTRUE:main_colorPurple collabTRUE:main_colorRed
## 1.328049 1.342645
## collabTRUE:main_colorWhite collabTRUE:main_colorYellow
## 2.193171 1.373942
## Linear mixed model fit by REML ['lmerMod']
## Formula: log_resale_price ~ (1 | brand) + release_month + release_dow +
## wmns + collab + retro + kids + main_color + bin_wants + retail_price
## Data: sneakers
##
## REML criterion at convergence: 7286.3
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -5.8785 -0.5723 -0.0271 0.4832 8.7730
##
## Random effects:
## Groups Name Variance Std.Dev.
## brand (Intercept) 0.008084 0.08991
## Residual 0.184705 0.42977
## Number of obs: 6164, groups: brand, 12
##
## Fixed effects:
## Estimate Std. Error t value
## (Intercept) 4.0736543 0.0476133 85.557
## release_monthFeb 0.0213021 0.0297161 0.717
## release_monthMar 0.0547314 0.0296867 1.844
## release_monthApr 0.0192001 0.0301579 0.637
## release_monthMay -0.0329924 0.0314958 -1.048
## release_monthJun 0.0284872 0.0304721 0.935
## release_monthJul -0.0028104 0.0310392 -0.091
## release_monthAug 0.0217993 0.0307140 0.710
## release_monthSep 0.0668168 0.0317158 2.107
## release_monthOct 0.0417167 0.0324003 1.288
## release_monthNov 0.0916576 0.0306865 2.987
## release_monthDec 0.0709083 0.0310721 2.282
## release_dowMonday 0.1779192 0.0357938 4.971
## release_dowTuesday 0.0684542 0.0332779 2.057
## release_dowWednesday 0.0629387 0.0306676 2.052
## release_dowThursday 0.0472275 0.0269842 1.750
## release_dowFriday 0.0448672 0.0282779 1.587
## release_dowSaturday 0.0791715 0.0276000 2.869
## wmnsTRUE 0.0336349 0.0164518 2.044
## collabTRUE 0.2580805 0.0193282 13.353
## retroTRUE 0.1511631 0.0329742 4.584
## kidsTRUE -0.0610148 0.0228466 -2.671
## main_colorBlue 0.0022677 0.0165131 0.137
## main_colorBrown -0.0620913 0.0355652 -1.746
## main_colorGray -0.0089108 0.0280363 -0.318
## main_colorGreen 0.0147141 0.0376136 0.391
## main_colorMulti-Color 0.0616943 0.0401102 1.538
## main_colorOrange -0.0289575 0.0356443 -0.812
## main_colorPurple -0.0233059 0.0353917 -0.659
## main_colorRed 0.0205475 0.0234084 0.878
## main_colorWhite 0.0272792 0.0145749 1.872
## main_colorYellow -0.1027996 0.0658986 -1.560
## bin_wantsSomewhat Popular 0.0117084 0.0170289 0.688
## bin_wantsVery Popular 0.1137923 0.0175774 6.474
## bin_wantsExtremely Popular 0.3047638 0.0191639 15.903
## retail_price 0.0033335 0.0001014 32.859